A blockchain-based private data sharing system

Abstract

The application discloses a kind of privacy data sharing systems based on block chain, comprising: ciphertext construction module, for obtaining and pre-processing privacy data to be shared;State commitment module is used to perform hash mapping and state coding, constructs shared state matrix and Merkle hash tree, and matrix root value is submitted to block chain node;Operator construction module is used to carry out Laplace spectrum reforming, state locking is carried out to effective ciphertext block and interpolation reconstruction is carried out;Re-encryption module is used to drive re-encryption, and the ciphertext component that completes driving is implemented conditional decoupling and directional reconstruction and executes constraint binding, solidifies limited solvable relationship;Declassification output module is used to execute path backtracking and check to target ciphertext block set and executes failure isolation, and outputs constrained sharing result;State update module is used to carry out state migration and root value update.The application can realize the controlled sharing of privacy data under the state commitment constraint on chain, and improve the credibility of data sharing.

Description

Technical Field

[0001] This invention relates to the field of data sharing technology, and in particular to a blockchain-based privacy data sharing system. Background Technology

[0002] With the continuous growth in demand for data sharing, a large amount of data resources with circulation value have been generated in scenarios such as government collaboration, medical services, financial risk control, and industrial internet. This type of data is usually both sensitive and relevant. In the process of sharing across entities, it is necessary to ensure sharing efficiency while preventing unauthorized access, illegal copying, or dissemination of the original data outside the authorized scope. Therefore, privacy data sharing mechanisms have gradually become an important research direction in the field of data security utilization.

[0003] In existing technologies, one type of solution typically uses a centralized server to complete identity authentication, access control, and data distribution. While this type of solution can achieve basic permission management, the shared rules, authorization process, and access records are mainly concentrated in a single management node. Once the central node is tampered with, bypassed, or becomes invalid, it is difficult to continuously verify the authenticity of data call paths, authorization boundaries, and shared results, making it difficult to meet the requirements of traceability and tamper resistance in a multi-entity collaborative environment.

[0004] Another type of approach typically employs traditional encryption mechanisms to protect sensitive data and discloses the data content to the target subject through key distribution or session decryption during sharing. While this approach can reduce the risk of plaintext exposure, in scenarios involving multi-round sharing, fine-grained sharing, and dynamic authorization, it often suffers from problems such as statically fixed sharing boundaries and difficulty in timely invalidation of sharing capabilities after changes in authorization status, making it difficult to achieve continuous constraints on the scope and results of sharing.

[0005] Another approach attempts to introduce blockchain into the privacy data sharing process, recording sharing behavior through on-chain evidence storage or enforcing authorization through smart contracts. While this approach improves the transparency and immutability of sharing records, in most cases the blockchain only participates in recording and auditing, failing to directly affect ciphertext conversion, decryption authorization, and output scope control. This results in a weak integration between on-chain state and off-chain ciphertext sharing processes, making it difficult to balance sharing credibility, privacy protection, and controllable sharing boundaries.

[0006] Therefore, how to provide a blockchain-based privacy data sharing system is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0007] One objective of this invention is to propose a blockchain-based privacy data sharing system. This invention fully utilizes technologies such as shared state matrix construction, Laplace spectrum renormalization, threshold proxy re-encryption, Merkel authentication path verification, and interval tree constraint positioning. It describes in detail the process of ciphertext sharing, decryption output, and state update of privacy data under on-chain state commitment constraints, and has the advantages of traceable sharing process, controllable sharing boundary, and strong privacy data protection.

[0008] A blockchain-based privacy data sharing system according to an embodiment of the present invention includes: The ciphertext construction module is used to acquire the private data to be shared and preprocess it to form a source ciphertext block set; The state commitment module is used to perform hash mapping and state encoding on the source ciphertext block set, construct a shared state matrix and Merkle hash tree, and submit the matrix root value to the blockchain node to complete consensus registration, forming an on-chain state commitment. The operator construction module is used to perform Laplace spectrum renormalization on the shared state matrix based on on-chain state commitment, lock the state of valid ciphertext blocks in combination with the authentication path, and use a threshold proxy re-encryption mechanism to interpolate and reconstruct the re-encryption key fragments to construct the proxy re-encryption operator. The re-encryption module is used to drive threshold re-encryption of the source ciphertext block set using the proxy re-encryption operator, implement conditional decoupling and directional reconstruction of the ciphertext components after the drive is completed, and bind them with the on-chain state commitment execution constraints to solidify the restricted solvable relationship of the target ciphertext block set. The decryption output module is used to perform Merkel authentication path backtracking and state root value verification on the target ciphertext block set, implement conditional decryption and range truncation based on the verification results, and perform failure isolation on the ciphertext components that fail the verification, and output constrained sharing results. The state update module is used to perform state transitions and root value updates on the shared state matrix based on the constrained shared results, and submit them to the blockchain node to complete consensus registration.

[0009] Optionally, the ciphertext construction module includes: Receive privacy data to be shared, perform structural organization and continuous reorganization on the privacy data to be shared, and form a continuous data payload; A Rabin fingerprint-based rolling scan is performed on the continuous data payload. The block boundaries are located according to the fingerprint values ​​and preset boundary determination rules. The continuous data payload is then divided into content-defined blocks according to the block boundaries to form the original data block set. Block identifier construction is performed on the original data block set. The digest value and position number of each original data block are combined into a block identifier. The original data blocks are then encapsulated at the block level according to the block identifier to form a data block set to be encrypted. Based on the block identifier, a key derivation operation is performed on the initial key to obtain the block-level key corresponding to the set of data blocks to be encrypted. The block-level key is then used to perform block encryption on the set of data blocks to be encrypted, forming the source ciphertext block set.

[0010] Optionally, the state commitment module includes: Hash calculations are performed on each source ciphertext block in the source ciphertext block set to obtain the block hash value corresponding to each source ciphertext block. The block hash values ​​are then arranged in order according to the block order of each source ciphertext block to form a block hash value sequence. According to the preset state coding rules, state quantization is performed on each source ciphertext block. The quantization results are binary encoded according to a fixed bit width and arranged in block order to form a state coding sequence. The preset state coding rules are used to perform interval partitioning and position mapping on each source ciphertext block according to the block hash value and block order corresponding to each source ciphertext block, and combine the interval partitioning results and position mapping results into a fixed-length binary coding value. The block hash value sequence and the state coding sequence are paired one by one according to the corresponding block order. The first and last parts of each pairing result are spliced ​​together and the position is marked to form a state mapping symbol sequence. The state mapping symbol sequence is then grouped and interleaved according to the preset matrix dimension to construct a shared state matrix. Perform hash compression on each row of the shared state matrix to obtain a row compressed value sequence. Perform pairwise concatenation and hash compression again on adjacent compressed values ​​in the row compressed value sequence to form a column compressed value sequence. Then, perform hierarchical hash folding on the column compressed value sequence to obtain the matrix root value of the shared state matrix. The matrix root value is broadcast to the blockchain nodes, and the blockchain nodes use the PBFT consensus algorithm to sequentially perform pre-preparation, preparation, commit, and block confirmation on the matrix root value, forming an on-chain state commitment.

[0011] Optionally, the construction of the shared state matrix includes: Each block hash value in the block hash value sequence is paired with the corresponding state code in the state code sequence according to the same block order to form a pair code group; The block hash value and state code in each paired code group are concatenated end to end, and the position identifier of the corresponding block order is embedded in the concatenation result to form a state mapping code symbol sequence; The state mapping symbol sequence is divided into windows according to the preset matrix dimensions. The state mapping symbols in each window are divided into continuous symbol groups, and the continuous symbol groups are length-normalized to form symbol groups to be arranged. The Z-shaped scanning algorithm is used to perform matrix path arrangement of the code groups to be arranged. Each code group to be arranged is alternately unfolded along the diagonal direction of the shared state matrix to form a matrix arrangement path. According to the matrix arrangement path, each code group to be arranged is embedded into the corresponding position of the shared state matrix in sequence, and adjacent state mapping codes after embedding are subjected to adjacency transposition and order reordering to form the shared state matrix.

[0012] Optionally, the operator construction module includes: Based on on-chain state commitments, Laplace matrix construction is performed on the shared state matrix, and the main spectrum extraction of the Laplace matrix is ​​performed using the Lanczos iterative algorithm. The eigenvectors corresponding to the main spectral components are projected onto the shared state matrix in block order. Threshold compression and connected component stripping are performed on the projection results to form a locked state submatrix. Based on the locked state submatrix, the Merkle hash path corresponding to the source ciphertext block set is backtracked from the leaf node to the root node. The hash value of each path node is bound to the corresponding state component in the locked state submatrix. The binding result is then subjected to path truncation and branch filtering to form a valid ciphertext block group corresponding to the locked state submatrix. Threshold aggregation and interpolation restoration are performed on the re-encryption key fragments corresponding to the valid ciphertext block groups, and block order registration is performed on the restoration results to form a key component sequence, specifically including: Based on the block order corresponding to each valid ciphertext block in the valid ciphertext block group, candidate fragments corresponding to the same block order are extracted from the re-encryption key fragments, and the candidate fragments are grouped and arranged according to the block order to form a fragment candidate group. Threshold screening is performed on the re-encryption key fragments in each fragment candidate group. The re-encryption key fragments with a fragment number reaching the threshold are bundled into fragment combinations. Coordinate binding is performed on the fragment index value and fragment value in each fragment combination to form an interpolation fragment group. Construct a Lagrange basis function for the fragment index value in each interpolation fragment group, perform multiplicative expansion and additive convergence on the fragment values ​​corresponding to each Lagrange basis function, and restore the re-encryption key component corresponding to each fragment combination; Perform block order back-labeling and sequence registration on each re-encrypted key component obtained from the reconstruction, and locate the re-encrypted key components with the same block order to the corresponding sequence position to form a key component sequence that corresponds one-to-one with the effective ciphertext block group. The state components in the locked state submatrix and the re-encryption key components in the key component sequence are assembled diagonally in the corresponding block order. The assembly result is then subjected to parameter compression and position fixation to form an operator unit sequence. The operator unit sequences are then spliced ​​together along the block order to form a proxy re-encryption operator.

[0013] Optionally, the formation of the locked state submatrix includes: Based on the block order adjacency relationship and corresponding state value of each matrix element in the shared state matrix, an adjacency matrix is ​​constructed, and the degree values ​​of the adjacency weights of each matrix element are aggregated to form a Laplace matrix corresponding to the adjacency matrix. The Laplacian matrix is ​​subjected to Lanczos recursion along the Krylov subspace. Three recursive constraints are applied to the basis vector set generated by the recursion to form a tridiagonal projection matrix corresponding to the Laplacian matrix. The principal eigenvalue interval is identified on the tridiagonal projection matrix, and the spectral components within the preset principal eigenvalue interval are projected along the block order direction to the corresponding block positions of the shared state matrix to form the principal spectral projection layer. The spectral gap threshold screening is performed on the main spectrum projection layer, and the continuous projection components are retained and the discrete projection components are stripped according to the block order adjacency relationship in the shared state matrix to form a locked state submatrix.

[0014] Optionally, the re-encryption module includes: The proxy re-encryption operator is expanded into a sequence of driving operator units in block order. Threshold Lagrange weighted driving is applied to each ciphertext component in the source ciphertext block set, so that each ciphertext component completes component-level re-encryption activation under the action of the corresponding driving operator unit, forming a candidate re-encryption component cluster. Conditional decoupling and adjoint term culling are performed on the candidate re-encryption component clusters. Non-target coupling components are stripped along the locking state constraints in the proxy re-encryption operator, and intra-block reorganization is performed on the retained components to form a purified re-encryption component cluster. The cleaned and re-encrypted component clusters are folded and repositioned according to block order. The cleaned and re-encrypted components under the same block order are pressed into the corresponding reconstructed ciphertext units, and the reconstructed ciphertext units are arranged into the target ciphertext block set along the block order. The target ciphertext block set is attached to the on-chain state commitment by performing root value locking, and state constraint fixation and block order binding are applied to each reconstructed ciphertext unit, so that the solvable range of the target ciphertext block set is limited to the corresponding on-chain state commitment, thus solidifying the limited solvable relationship of the target ciphertext block set.

[0015] Optionally, the formation of the candidate re-encryption component cluster includes: The proxy re-encryption operator is expanded along the block order into driving operator units corresponding to the source ciphertext block set. The state components and re-encryption parameters in each driving operator unit are arranged within the block to form driving primitives corresponding to each source ciphertext block. Threshold clustering is performed on the re-encryption key components corresponding to each driving primitive. Lagrange basis functions are used to construct interpolation weights for the re-encryption key components that reach the threshold number, and the interpolation weights are pressed onto the corresponding driving primitives to form a threshold driving kernel. The threshold-driven kernel is invoked to perform component-level re-encryption multiplication on each ciphertext component in the source ciphertext block set. The accompaniment term suppression and effective component traction are performed on the multiplied ciphertext components, so that the retained effective components enter the re-encryption activation state and form candidate re-encryption components. For each candidate re-encryption component, perform block order convergence and same-position fixed arrangement, and arrange the candidate re-encryption components under the same block order into corresponding candidate re-encryption component clusters according to the component position relationship.

[0016] Optionally, the decryption output module includes: Based on each target ciphertext block in the target ciphertext block set, a layer-by-layer back-hash folding is performed from the leaf node to the root node along the corresponding Merkel authentication path. The hash values ​​of the path nodes are recursively reorganized in hierarchical order to form the path root value group corresponding to each target ciphertext block. A block order parity check is performed on the path root value group and the state root value in the on-chain state commitment. The target ciphertext block that passes the check is spread into a decryption permission domain, and the target ciphertext block that fails the check is spread into an invalidation isolation domain. The decryption parameters are reverse-engineered and gated decryption is performed on the target ciphertext block within the decryption permission domain. The decryption results are then reordered to form a candidate plaintext block group, specifically including: Each target ciphertext block within the decryption license domain is expanded into a sequence of ciphertext components in block order, and each ciphertext component and its corresponding decryption kernel are loaded in the same block order to form a decryption driver group. Gated decryption expansion is performed on each decryption driving group, and Gaussian elimination is used to perform inverse solution of the component transformation relationship in the decryption kernel to obtain the inverse decryption parameter group corresponding to each ciphertext component; The inverse decryption parameter set is loaded into each ciphertext component along the corresponding component position, and the gated inverse decryption and component domain release are performed on each ciphertext component to form a candidate plaintext component set corresponding to each target ciphertext block. For each candidate plaintext component group, perform segment reassembly and block order convergence according to the original component position relationship to form a candidate plaintext block group; The interval tree constraint positioning method is used to perform output interval determination for candidate plaintext block groups. Candidate plaintext components falling into the preset output interval are released into restricted plaintext block sets along the block order, and target ciphertext blocks in the failure isolation domain are isolated and shielded, and constrained sharing results are output.

[0017] Optionally, the state update module includes: The constrained shared results are pressed into the shared state matrix in block order, and the corresponding matrix elements in the shared state matrix are registered to form state backfeed terms. The offset residual expansion is performed on the state recharge term, and the recursive least squares method is used to recursively converge the residual offset between the constrained shared result and the corresponding matrix element to form the state transition increment matrix. The state transition increment matrix is ​​embedded into the shared state matrix, and incremental replacement and block order constraint convergence are performed on the corresponding matrix elements in the shared state matrix to form the updated shared state matrix. Perform row-wise hash folding and column-wise hash folding on the updated shared state matrix to form the updated matrix root value, and submit the updated matrix root value to the blockchain node to complete consensus registration, forming the updated on-chain state commitment.

[0018] The beneficial effects of this invention are: First, this invention introduces on-chain state commitments into the privacy data sharing process by constructing a shared state matrix and submitting the matrix root value to the blockchain node to complete consensus registration. This eliminates the reliance on centralized nodes for authorization and recording, making state changes and outputs in the sharing process traceable and tamper-resistant, thus enhancing the credibility of data sharing in multi-entity collaborative scenarios.

[0019] Secondly, this invention links on-chain state commitments with off-chain ciphertext block flow processes through Laplace spectrum re-normalization, threshold proxy re-encryption, and Merkel authentication path verification. This enables the effective ciphertext block locking, re-encryption operator construction, and decryption output control to form a continuous technical chain, avoiding the problem of the authorization record being disconnected from the actual ciphertext sharing behavior in the prior art.

[0020] Finally, this invention enables the shared results to be released in a controlled manner within a preset range through conditional decryption, interval constraint output, and state transition update of the shared state matrix, and continues to apply the shared results of this round to the subsequent state commitment process, thus taking into account privacy protection, controllability of sharing, and stability of system operation. Attached Figure Description

[0021] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a block diagram of a blockchain-based privacy data sharing system proposed in this invention. Figure 2 This is a flowchart illustrating the state commitment construction process for a blockchain-based privacy data sharing system proposed in this invention. Figure 3 This is a flowchart illustrating the construction of a proxy re-encryption operator for a blockchain-based privacy data sharing system proposed in this invention. Detailed Implementation

[0022] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0023] refer to Figures 1-3 A blockchain-based privacy data sharing system includes: The ciphertext construction module is used to acquire the private data to be shared and preprocess it to form a source ciphertext block set; The state commitment module is used to perform hash mapping and state encoding on the source ciphertext block set, construct a shared state matrix and Merkle hash tree, and submit the matrix root value to the blockchain node to complete consensus registration, forming an on-chain state commitment. The operator construction module is used to perform Laplace spectrum renormalization on the shared state matrix based on on-chain state commitment, lock the state of valid ciphertext blocks in combination with the authentication path, and use a threshold proxy re-encryption mechanism to interpolate and reconstruct the re-encryption key fragments to construct the proxy re-encryption operator. The re-encryption module is used to drive threshold re-encryption of the source ciphertext block set using the proxy re-encryption operator, implement conditional decoupling and directional reconstruction of the ciphertext components after the drive is completed, and bind them with the on-chain state commitment execution constraints to solidify the restricted solvable relationship of the target ciphertext block set. The decryption output module is used to perform Merkel authentication path backtracking and state root value verification on the target ciphertext block set, implement conditional decryption and range truncation based on the verification results, and perform failure isolation on the ciphertext components that fail the verification, and output constrained sharing results. The state update module is used to perform state transitions and root value updates on the shared state matrix based on the constrained shared results, and submit them to the blockchain node to complete consensus registration.

[0024] In this embodiment, the ciphertext construction module includes: Receive privacy data to be shared, perform structural organization and continuous reorganization on the privacy data to be shared, and form a continuous data payload; A Rabin fingerprint-based rolling scan is performed on the continuous data payload. The block boundaries are located according to the fingerprint values ​​and preset boundary determination rules. The continuous data payload is then divided into content-defined blocks according to the block boundaries to form the original data block set. Block identifier construction is performed on the original data block set. The digest value and position number of each original data block are combined into a block identifier. The original data blocks are then encapsulated at the block level according to the block identifier to form a data block set to be encrypted. Based on the block identifier, a key derivation operation is performed on the initial key to obtain the block-level key corresponding to the set of data blocks to be encrypted. The block-level key is then used to perform block encryption on the set of data blocks to be encrypted, forming the source ciphertext block set.

[0025] In this embodiment, the state commitment module includes: Hash calculations are performed on each source ciphertext block in the source ciphertext block set to obtain the block hash value corresponding to each source ciphertext block. These block hash values ​​are then arranged sequentially according to the block order of the source ciphertext blocks to form a block hash value sequence, specifically including: The length of each source ciphertext block in the source ciphertext block set is checked according to the block order. Source ciphertext blocks whose length is less than the preset length are padded at the end, and source ciphertext blocks whose length exceeds the preset length are segmented into fixed-length block data units. Perform byte rearrangement and segment replacement on each fixed-length block data unit, and divide the rearranged byte stream into multiple data segments according to the preset segment length to form a segmented data group corresponding to each fixed-length block data unit; The SHA-256 hash algorithm is used to perform iterative compression operations on each segment of data group. Each data segment is sent to the compression round function for round update to obtain the block hash value corresponding to each fixed-length block data unit. Sliding merge is performed on the block hash values ​​corresponding to adjacent block sequences according to the preset window length. The block hash values ​​in the same window are concatenated sequentially and then hash compression is performed again to obtain the window hash segment. According to the block order range covered by each window hash segment, the window hash segments are back-indexed and expanded sequentially, and the expanded window hash segments are reorganized into a block hash value sequence according to the block order. According to the preset state coding rules, state quantization is performed on each source ciphertext block. The quantization results are binary encoded according to a fixed bit width and arranged in block order to form a state coding sequence. The preset state coding rules are used to perform interval partitioning and position mapping on each source ciphertext block according to the block hash value and block order corresponding to each source ciphertext block, and combine the interval partitioning results and position mapping results into a fixed-length binary coding value. The block hash value sequence and the state coding sequence are paired one by one according to the corresponding block order. The first and last parts of each pairing result are spliced ​​together and the position is marked to form a state mapping symbol sequence. The state mapping symbol sequence is then grouped and interleaved according to the preset matrix dimension to construct a shared state matrix. Perform hash compression on each row of the shared state matrix to obtain a row compressed value sequence. Perform pairwise concatenation and hash compression again on adjacent compressed values ​​in the row compressed value sequence to form a column compressed value sequence. Then, perform hierarchical hash folding on the column compressed value sequence to obtain the matrix root value of the shared state matrix. The matrix root value is broadcast to the blockchain nodes, and the blockchain nodes use the PBFT consensus algorithm to sequentially perform pre-preparation, preparation, commit, and block confirmation on the matrix root value, forming an on-chain state commitment.

[0026] In this embodiment, the construction of the shared state matrix includes: Each block hash value in the block hash value sequence is paired with the corresponding state code in the state code sequence according to the same block order to form a pair code group; The block hash value and state code in each paired code group are concatenated end to end, and the position identifier of the corresponding block order is embedded in the concatenation result to form a state mapping code symbol sequence; The state mapping symbol sequence is divided into windows according to the preset matrix dimensions. The state mapping symbols in each window are divided into continuous symbol groups, and the continuous symbol groups are length-normalized to form symbol groups to be arranged. The Z-shaped scanning algorithm is used to perform matrix path arrangement of the code groups to be arranged. Each code group to be arranged is alternately unfolded along the diagonal direction of the shared state matrix to form a matrix arrangement path. According to the matrix arrangement path, each code group to be arranged is embedded into the corresponding position of the shared state matrix in sequence, and adjacent state mapping codes after embedding are subjected to adjacency transposition and order reordering to form the shared state matrix.

[0027] In this embodiment, the operator construction module includes: Based on on-chain state commitments, Laplace matrix construction is performed on the shared state matrix, and the main spectrum extraction of the Laplace matrix is ​​performed using the Lanczos iterative algorithm. The eigenvectors corresponding to the main spectral components are projected onto the shared state matrix in block order. Threshold compression and connected component stripping are performed on the projection results to form a locked state submatrix. Based on the locked state submatrix, the Merkle hash path corresponding to the source ciphertext block set is backtracked from the leaf node to the root node. The hash value of each path node is bound to the corresponding state component in the locked state submatrix. The binding result is then subjected to path truncation and branch filtering to form a valid ciphertext block group corresponding to the locked state submatrix. Threshold aggregation and interpolation restoration are performed on the re-encryption key fragments corresponding to the valid ciphertext block groups, and block order registration is performed on the restoration results to form a key component sequence, specifically including: Based on the block order corresponding to each valid ciphertext block in the valid ciphertext block group, candidate fragments corresponding to the same block order are extracted from the re-encryption key fragments, and the candidate fragments are grouped and arranged according to the block order to form a fragment candidate group. Threshold screening is performed on the re-encryption key fragments in each fragment candidate group. The re-encryption key fragments with a fragment number reaching the threshold are bundled into fragment combinations. Coordinate binding is performed on the fragment index value and fragment value in each fragment combination to form an interpolation fragment group. Construct a Lagrange basis function for the fragment index value in each interpolation fragment group, perform multiplicative expansion and additive convergence on the fragment values ​​corresponding to each Lagrange basis function, and restore the re-encryption key component corresponding to each fragment combination; Perform block order back-labeling and sequence registration on each re-encrypted key component obtained from the reconstruction, and locate the re-encrypted key components with the same block order to the corresponding sequence position to form a key component sequence that corresponds one-to-one with the effective ciphertext block group. The state components in the locked state submatrix and the re-encryption key components in the key component sequence are assembled diagonally in the corresponding block order. The assembly result is then subjected to parameter compression and position fixation to form an operator unit sequence. The operator unit sequences are then spliced ​​together along the block order to form a proxy re-encryption operator.

[0028] In this embodiment, the formation of the locked state sub-matrix includes: Based on the block order adjacency relationship and corresponding state value of each matrix element in the shared state matrix, an adjacency matrix is ​​constructed, and the degree values ​​of the adjacency weights of each matrix element are aggregated to form a Laplace matrix corresponding to the adjacency matrix. The Laplacian matrix is ​​subjected to Lanczos recursion along the Krylov subspace. Three recursive constraints are applied to the basis vector set generated by the recursion to form a tridiagonal projection matrix corresponding to the Laplacian matrix. The principal eigenvalue interval is identified on the tridiagonal projection matrix, and the spectral components within the preset principal eigenvalue interval are projected along the block order direction to the corresponding block positions of the shared state matrix to form the principal spectral projection layer. The spectral gap threshold screening is performed on the main spectrum projection layer, and the continuous projection components are retained and the discrete projection components are stripped according to the block order adjacency relationship in the shared state matrix to form a locked state submatrix.

[0029] In this embodiment, the re-encryption module includes: The proxy re-encryption operator is expanded into a sequence of driving operator units in block order. Threshold Lagrange weighted driving is applied to each ciphertext component in the source ciphertext block set, so that each ciphertext component completes component-level re-encryption activation under the action of the corresponding driving operator unit, forming a candidate re-encryption component cluster. Conditional decoupling and adjoint term culling are performed on the candidate re-encryption component clusters. Non-target coupled components are stripped along the locked state constraints in the proxy re-encryption operator, and intra-block reorganization is performed on the retained components to form a purified re-encryption component cluster, specifically including: Expand the candidate re-encryption component clusters into component matrices according to block order, and press the locking state constraints in the proxy re-encryption operator onto the component matrices along the corresponding block order to form a constraint component matrix; Perform constraint projection on the constraint component matrix to compress the candidate re-encryption component that satisfies the locked state constraint into the main component submatrix, and pull out the non-target candidate re-encryption component coupled with the main component submatrix into the adjoint component submatrix to form the coupled component matrix. The coupling components between the principal component submatrix and the adjoint component submatrix in the coupling component matrix are subjected to Givens rotation elimination to reduce the coupling components to below the preset hidden component elimination threshold, and the corresponding adjoint components are stripped from the coupling component matrix to form a clean component matrix. Perform intra-block re-erase of the cleaned component matrix in block order, and embed the retained cleaned and re-encrypted components back into the corresponding block positions along the preset component position relationship to form a cleaned and re-encrypted component cluster. The cleaned and re-encrypted component clusters are folded and repositioned according to block order. The cleaned and re-encrypted components under the same block order are pressed into the corresponding reconstructed ciphertext units, and the reconstructed ciphertext units are arranged into the target ciphertext block set along the block order. The target ciphertext block set is attached to the on-chain state commitment by performing root value locking, and state constraint fixation and block order binding are applied to each reconstructed ciphertext unit, so that the solvable range of the target ciphertext block set is limited to the corresponding on-chain state commitment, thus solidifying the limited solvable relationship of the target ciphertext block set.

[0030] In this embodiment, the formation of the candidate re-encryption component cluster includes: The proxy re-encryption operator is expanded along the block order into driving operator units corresponding to the source ciphertext block set. The state components and re-encryption parameters in each driving operator unit are arranged within the block to form driving primitives corresponding to each source ciphertext block. Threshold clustering is performed on the re-encryption key components corresponding to each driving primitive. Lagrange basis functions are used to construct interpolation weights for the re-encryption key components that reach the threshold number, and the interpolation weights are pressed onto the corresponding driving primitives to form a threshold driving kernel. The threshold-driven kernel is invoked to perform component-level re-encryption multiplication on each ciphertext component in the source ciphertext block set. The accompaniment term suppression and effective component traction are performed on the multiplied ciphertext components, so that the retained effective components enter the re-encryption activation state and form candidate re-encryption components. For each candidate re-encryption component, perform block order convergence and same-position fixed arrangement, and arrange the candidate re-encryption components under the same block order into corresponding candidate re-encryption component clusters according to the component position relationship.

[0031] In this embodiment, the decryption output module includes: Based on each target ciphertext block in the target ciphertext block set, a layer-by-layer back-hash folding is performed from the leaf node to the root node along the corresponding Merkel authentication path. The hash values ​​of the path nodes are recursively reorganized in hierarchical order to form the path root value group corresponding to each target ciphertext block. A block order parity check is performed on the path root value group and the state root value in the on-chain state commitment. The target ciphertext block that passes the check is spread into a decryption permission domain, and the target ciphertext block that fails the check is spread into an invalidation isolation domain. The decryption parameters are reverse-engineered and gated decryption is performed on the target ciphertext block within the decryption permission domain. The decryption results are then reordered to form a candidate plaintext block group, specifically including: Each target ciphertext block within the decryption license domain is expanded into a sequence of ciphertext components in block order, and each ciphertext component and its corresponding decryption kernel are loaded in the same block order to form a decryption driver group. Gated decryption expansion is performed on each decryption driving group, and Gaussian elimination is used to perform inverse solution of the component transformation relationship in the decryption kernel to obtain the inverse decryption parameter group corresponding to each ciphertext component; The inverse decryption parameter set is loaded into each ciphertext component along the corresponding component position, and the gated inverse decryption and component domain release are performed on each ciphertext component to form a candidate plaintext component set corresponding to each target ciphertext block. For each candidate plaintext component group, perform segment reassembly and block order convergence according to the original component position relationship to form a candidate plaintext block group; The interval tree constraint positioning method is used to perform output interval determination for candidate plaintext block groups. Candidate plaintext components falling into the preset output interval are released into restricted plaintext block sets along the block order, and target ciphertext blocks in the failure isolation domain are isolated and shielded, and constrained sharing results are output.

[0032] In this embodiment, the status update module includes: The constrained shared results are pressed into the shared state matrix in block order, and the corresponding matrix elements in the shared state matrix are registered to form state backfeed terms. The offset residual expansion is performed on the state recharge term, and the recursive least squares method is used to recursively converge the residual offset between the constrained shared result and the corresponding matrix element to form the state transition increment matrix. The state transition increment matrix is ​​embedded into the shared state matrix, and incremental replacement and block order constraint convergence are performed on the corresponding matrix elements in the shared state matrix to form the updated shared state matrix. Perform row-wise hash folding and column-wise hash folding on the updated shared state matrix to form the updated matrix root value, and submit the updated matrix root value to the blockchain node to complete consensus registration, forming the updated on-chain state commitment.

[0033] Example 1: To verify the feasibility of this invention in practice, it was applied to a cross-organizational privacy data collaborative verification scenario. In this scenario, multiple business nodes respectively store identity verification data, behavior record data, and historical authorization data. Each node needs to provide constrained data sharing capabilities to upper-layer collaborative analysis nodes without directly exposing the original data content. In existing methods, although sharing requests are registered, there is often a lack of close linkage between on-chain records and off-chain encrypted data flow. This results in the old sharing capabilities continuing to be effective after the authorization status changes. At the same time, the data granularity returned by different nodes is inconsistent, which can easily lead to uncontrolled sharing boundaries, output content exceeding the scope, and difficulties in subsequent traceability.

[0034] In this embodiment, the privacy data to be shared is first preprocessed and initially encrypted in blocks to form a source ciphertext block set. Then, the source ciphertext block set is hashed and state-encoded to construct a shared state matrix. The root value of the matrix is ​​submitted to the blockchain node to complete consensus registration, forming an on-chain state commitment. When a collaborative analysis node initiates a sharing request, the system performs Laplace spectrum reshaping based on the on-chain state commitment and the shared state matrix. It then locks valid ciphertext blocks using a Merkel authentication path and constructs a proxy re-encryption operator through a threshold proxy re-encryption mechanism to perform constrained re-encryption driving and conditional decoupling on the target ciphertext block set. During the decryption output phase, decryption permission is only granted to target ciphertext blocks that pass the state root value verification. The range of releaseable components is limited using an interval tree constraint positioning method, ultimately outputting a constrained sharing result. After sharing is complete, the output result is fed back into the shared state matrix. A recursive least squares method is used to form a state transition increment matrix, updating the matrix root value and re-completing the on-chain consensus registration, thus forming a continuous closed-loop sharing control process.

[0035] In a single continuous run, the system processed 12,800 source ciphertext blocks, of which 9,460 were valid. The remaining ciphertext blocks were isolated and masked due to failing the state root value check or being outside the permitted range. The average time to generate on-chain state commitments was 0.83 seconds, the average construction time of the proxy re-encryption operator was 1.17 seconds, and the average time for a single round of complete sharing was 2.96 seconds.

[0036] Compared to the conventional on-chain registration plus off-chain decryption method, this invention demonstrates more stable sharing control capabilities under continuous multi-round sharing conditions. In the conventional method, some old ciphertext can still participate in subsequent outputs after the authorization state changes, with an out-of-range component release rate of 3.8%. In contrast, this invention, through state root value verification, permitted domain control, and interval constraint positioning, reduces the out-of-range component release rate to 0.2%. In 48 abnormal requests, the conventional method only intercepted 39, while this invention intercepted 46, increasing the abnormal request interception rate from 81.3% to 95.8%. Simultaneously, the matching rate between the candidate plaintext blocks generated by this invention and the preset allowed output range reached 98.9%, the block order recovery accuracy reached 99.2%, and the matrix root value consistency pass rate remained at 99.6%, indicating a stable linkage between on-chain state commitment and the off-chain ciphertext sharing process. Therefore, this invention can achieve trusted sharing without exposing original privacy data and can restrict sharing behavior within preset boundaries, providing reliable technical support for subsequent traceability and continuous updates.

[0037] Table 1. Comparison of the Implementation Effects of Constrained Privacy Data Sharing

[0038] As can be seen from Table 1, the present invention outperforms the conventional method in terms of on-chain state commitment generation efficiency, proxy re-encryption processing efficiency, abnormal request interception capability, and output range control accuracy. In particular, it shows significant improvements in the two indicators of out-of-range component release ratio and block order recovery accuracy. This indicates that the present invention can effectively solve the problems of insufficient linkage between on-chain records and off-chain sharing, difficulty in controlling sharing boundaries, and untimely failure of dynamic authorization in the prior art.

[0039] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A blockchain-based privacy data sharing system, characterized in that, include: The ciphertext construction module is used to acquire the private data to be shared and preprocess it to form a source ciphertext block set; The state commitment module is used to perform hash mapping and state encoding on the source ciphertext block set, construct a shared state matrix and Merkle hash tree, and submit the matrix root value to the blockchain node to complete consensus registration, forming an on-chain state commitment. The operator construction module is used to perform Laplace spectrum renormalization on the shared state matrix based on on-chain state commitment, lock the state of valid ciphertext blocks in combination with the authentication path, and use a threshold proxy re-encryption mechanism to interpolate and reconstruct the re-encryption key fragments to construct the proxy re-encryption operator. The re-encryption module is used to drive threshold re-encryption of the source ciphertext block set using the proxy re-encryption operator, implement conditional decoupling and directional reconstruction of the ciphertext components after the drive is completed, and bind them with the on-chain state commitment execution constraints to solidify the restricted solvable relationship of the target ciphertext block set. The decryption output module is used to perform Merkel authentication path backtracking and state root value verification on the target ciphertext block set, implement conditional decryption and range truncation based on the verification results, and perform failure isolation on the ciphertext components that fail the verification, and output constrained sharing results. The state update module is used to perform state transitions and root value updates on the shared state matrix based on the constrained shared results, and submit them to the blockchain node to complete consensus registration.

2. The blockchain-based privacy data sharing system according to claim 1, characterized in that, The ciphertext construction module includes: Receive privacy data to be shared, perform structural organization and continuous reorganization on the privacy data to be shared, and form a continuous data payload; A Rabin fingerprint-based rolling scan is performed on the continuous data payload. The block boundaries are located according to the fingerprint values ​​and preset boundary determination rules. The continuous data payload is then divided into content-defined blocks according to the block boundaries to form the original data block set. Block identifier construction is performed on the original data block set. The digest value and position number of each original data block are combined into a block identifier. The original data blocks are then encapsulated at the block level according to the block identifier to form a data block set to be encrypted. Based on the block identifier, a key derivation operation is performed on the initial key to obtain the block-level key corresponding to the set of data blocks to be encrypted. The block-level key is then used to perform block encryption on the set of data blocks to be encrypted, forming the source ciphertext block set.

3. The blockchain-based privacy data sharing system according to claim 1, characterized in that, The state commitment module includes: Hash calculations are performed on each source ciphertext block in the source ciphertext block set to obtain the block hash value corresponding to each source ciphertext block. The block hash values ​​are then arranged in order according to the block order of each source ciphertext block to form a block hash value sequence. According to the preset state coding rules, state quantization is performed on each source ciphertext block. The quantization results are binary encoded according to a fixed bit width and arranged in block order to form a state coding sequence. The preset state coding rules are used to perform interval partitioning and position mapping on each source ciphertext block according to the block hash value and block order corresponding to each source ciphertext block, and combine the interval partitioning results and position mapping results into a fixed-length binary coding value. The block hash value sequence and the state coding sequence are paired one by one according to the corresponding block order. The first and last parts of each pairing result are spliced ​​together and the position is marked to form a state mapping symbol sequence. The state mapping symbol sequence is then grouped and interleaved according to the preset matrix dimension to construct a shared state matrix. Perform hash compression on each row of the shared state matrix to obtain a row compressed value sequence. Perform pairwise concatenation and hash compression again on adjacent compressed values ​​in the row compressed value sequence to form a column compressed value sequence. Then, perform hierarchical hash folding on the column compressed value sequence to obtain the matrix root value of the shared state matrix. The matrix root value is broadcast to the blockchain nodes, and the blockchain nodes use the PBFT consensus algorithm to sequentially perform pre-preparation, preparation, commit, and block confirmation on the matrix root value, forming an on-chain state commitment.

4. A blockchain-based privacy data sharing system according to claim 3, characterized in that, The construction of the shared state matrix includes: Each block hash value in the block hash value sequence is paired with the corresponding state code in the state code sequence according to the same block order to form a pair code group; The block hash value and state code in each paired code group are concatenated end to end, and the position identifier of the corresponding block order is embedded in the concatenation result to form a state mapping code symbol sequence; The state mapping symbol sequence is divided into windows according to the preset matrix dimensions. The state mapping symbols in each window are divided into continuous symbol groups, and the continuous symbol groups are length-normalized to form symbol groups to be arranged. The Z-shaped scanning algorithm is used to perform matrix path arrangement of the code groups to be arranged. Each code group to be arranged is alternately unfolded along the diagonal direction of the shared state matrix to form a matrix arrangement path. According to the matrix arrangement path, each code group to be arranged is embedded into the corresponding position of the shared state matrix in sequence, and adjacent state mapping codes after embedding are subjected to adjacency transposition and order reordering to form the shared state matrix.

5. A blockchain-based privacy data sharing system according to claim 1, characterized in that, The operator construction module includes: Based on on-chain state commitments, Laplace matrix construction is performed on the shared state matrix, and the main spectrum extraction of the Laplace matrix is ​​performed using the Lanczos iterative algorithm. The eigenvectors corresponding to the main spectral components are projected onto the shared state matrix in block order. Threshold compression and connected component stripping are performed on the projection results to form a locked state submatrix. Based on the locked state submatrix, the Merkle hash path corresponding to the source ciphertext block set is backtracked from the leaf node to the root node. The hash value of each path node is bound to the corresponding state component in the locked state submatrix. The binding result is then subjected to path truncation and branch filtering to form a valid ciphertext block group corresponding to the locked state submatrix. Threshold aggregation and interpolation restoration are performed on the re-encryption key fragments corresponding to the valid ciphertext block groups, and block order registration is performed on the restoration results to form a key component sequence, specifically including: Based on the block order corresponding to each valid ciphertext block in the valid ciphertext block group, candidate fragments corresponding to the same block order are extracted from the re-encryption key fragments, and the candidate fragments are grouped and arranged according to the block order to form a fragment candidate group. Threshold screening is performed on the re-encryption key fragments in each fragment candidate group. The re-encryption key fragments with a fragment number reaching the threshold are bundled into fragment combinations. Coordinate binding is performed on the fragment index value and fragment value in each fragment combination to form an interpolation fragment group. Construct a Lagrange basis function for the fragment index value in each interpolation fragment group, perform multiplicative expansion and additive convergence on the fragment values ​​corresponding to each Lagrange basis function, and restore the re-encryption key component corresponding to each fragment combination; Perform block order back-labeling and sequence registration on each re-encrypted key component obtained from the reconstruction, and locate the re-encrypted key components with the same block order to the corresponding sequence position to form a key component sequence that corresponds one-to-one with the effective ciphertext block group. The state components in the locked state submatrix and the re-encryption key components in the key component sequence are assembled diagonally in the corresponding block order. The assembly result is then subjected to parameter compression and position fixation to form an operator unit sequence. The operator unit sequences are then spliced ​​together along the block order to form a proxy re-encryption operator.

6. A blockchain-based privacy data sharing system according to claim 5, characterized in that, The formation of the locked state sub-matrix includes: Based on the block order adjacency relationship and corresponding state value of each matrix element in the shared state matrix, an adjacency matrix is ​​constructed, and the degree values ​​of the adjacency weights of each matrix element are aggregated to form a Laplace matrix corresponding to the adjacency matrix. The Laplacian matrix is ​​subjected to Lanczos recursion along the Krylov subspace. Three recursive constraints are applied to the basis vector set generated by the recursion to form a tridiagonal projection matrix corresponding to the Laplacian matrix. The principal eigenvalue interval is identified on the tridiagonal projection matrix, and the spectral components within the preset principal eigenvalue interval are projected along the block order direction to the corresponding block positions of the shared state matrix to form the principal spectral projection layer. The spectral gap threshold screening is performed on the main spectrum projection layer, and the continuous projection components are retained and the discrete projection components are stripped according to the block order adjacency relationship in the shared state matrix to form a locked state submatrix.

7. A blockchain-based privacy data sharing system according to claim 1, characterized in that, The re-encryption module includes: The proxy re-encryption operator is expanded into a sequence of driving operator units in block order. Threshold Lagrange weighted driving is applied to each ciphertext component in the source ciphertext block set, so that each ciphertext component completes component-level re-encryption activation under the action of the corresponding driving operator unit, forming a candidate re-encryption component cluster. Conditional decoupling and adjoint term culling are performed on the candidate re-encryption component clusters. Non-target coupling components are stripped along the locking state constraints in the proxy re-encryption operator, and intra-block reorganization is performed on the retained components to form a purified re-encryption component cluster. The cleaned and re-encrypted component clusters are folded and repositioned according to block order. The cleaned and re-encrypted components under the same block order are pressed into the corresponding reconstructed ciphertext units, and the reconstructed ciphertext units are arranged into the target ciphertext block set along the block order. The target ciphertext block set is attached to the on-chain state commitment by performing root value locking, and state constraint fixation and block order binding are applied to each reconstructed ciphertext unit, so that the solvable range of the target ciphertext block set is limited to the corresponding on-chain state commitment, thus solidifying the limited solvable relationship of the target ciphertext block set.

8. A blockchain-based privacy data sharing system according to claim 7, characterized in that, The formation of the candidate re-encryption component cluster includes: The proxy re-encryption operator is expanded along the block order into driving operator units corresponding to the source ciphertext block set. The state components and re-encryption parameters in each driving operator unit are arranged within the block to form driving primitives corresponding to each source ciphertext block. Threshold clustering is performed on the re-encryption key components corresponding to each driving primitive. Lagrange basis functions are used to construct interpolation weights for the re-encryption key components that reach the threshold number, and the interpolation weights are pressed onto the corresponding driving primitives to form a threshold driving kernel. The threshold-driven kernel is invoked to perform component-level re-encryption multiplication on each ciphertext component in the source ciphertext block set. The accompaniment term suppression and effective component traction are performed on the multiplied ciphertext components, so that the retained effective components enter the re-encryption activation state and form candidate re-encryption components. For each candidate re-encryption component, perform block order convergence and same-position fixed arrangement, and arrange the candidate re-encryption components under the same block order into corresponding candidate re-encryption component clusters according to the component position relationship.

9. A blockchain-based privacy data sharing system according to claim 1, characterized in that, The decryption output module includes: Based on each target ciphertext block in the target ciphertext block set, a layer-by-layer back-hash folding is performed from the leaf node to the root node along the corresponding Merkel authentication path. The hash values ​​of the path nodes are recursively reorganized in hierarchical order to form the path root value group corresponding to each target ciphertext block. A block order parity check is performed on the path root value group and the state root value in the on-chain state commitment. The target ciphertext block that passes the check is spread into a decryption permission domain, and the target ciphertext block that fails the check is spread into an invalidation isolation domain. The decryption parameters are reverse-engineered and gated decryption is performed on the target ciphertext block within the decryption permission domain. The decryption results are then reordered to form a candidate plaintext block group, specifically including: Each target ciphertext block within the decryption license domain is expanded into a sequence of ciphertext components in block order, and each ciphertext component and its corresponding decryption kernel are loaded in the same block order to form a decryption driver group. Gated decryption expansion is performed on each decryption driving group, and Gaussian elimination is used to perform inverse solution of the component transformation relationship in the decryption kernel to obtain the inverse decryption parameter group corresponding to each ciphertext component; The inverse decryption parameter set is loaded into each ciphertext component along the corresponding component position, and the gated inverse decryption and component domain release are performed on each ciphertext component to form a candidate plaintext component set corresponding to each target ciphertext block. For each candidate plaintext component group, perform segment reassembly and block order convergence according to the original component position relationship to form a candidate plaintext block group; The interval tree constraint positioning method is used to perform output interval determination for candidate plaintext block groups. Candidate plaintext components falling into the preset output interval are released into restricted plaintext block sets along the block order, and target ciphertext blocks in the failure isolation domain are isolated and shielded, and constrained sharing results are output.

10. A blockchain-based privacy data sharing system according to claim 1, characterized in that, The status update module includes: The constrained shared results are pressed into the shared state matrix in block order, and the corresponding matrix elements in the shared state matrix are registered to form state backfeed terms. The offset residual expansion is performed on the state recharge term, and the recursive least squares method is used to recursively converge the residual offset between the constrained shared result and the corresponding matrix element to form the state transition increment matrix. The state transition increment matrix is ​​embedded into the shared state matrix, and incremental replacement and block order constraint convergence are performed on the corresponding matrix elements in the shared state matrix to form the updated shared state matrix. Perform row-wise hash folding and column-wise hash folding on the updated shared state matrix to form the updated matrix root value, and submit the updated matrix root value to the blockchain node to complete consensus registration, forming the updated on-chain state commitment.

Images